Download Effects of low flow on dwarf galaxias and their habitat in the Wairau

Effects of low flow on
dwarf galaxias and their
habitat in the Wairau River
Joe Hay
DOC Research & Development Series 309
Published by
Publishing Team
Department of Conservation
PO Box 10420, The Terrace
Wellington 6143, New Zealand
DOC Research & Development Series is a published record of scientific research carried out, or advice
given, by Department of Conservation staff or external contractors funded by DOC. It comprises reports
and short communications that are peer-reviewed.
Individual contributions to the series are first released on the departmental website in pdf form.
Hardcopy is printed, bound, and distributed at regular intervals. Titles are also listed in our catalogue on
the website, refer www.doc.govt.nz under Publications, then Science & technical.
© Copyright April 2009, New Zealand Department of Conservation
ISSN
ISSN
ISBN
ISBN
1176–8886 (hardcopy)
1177–9306 (web PDF)
978–0–478–14595–3 (hardcopy)
978–0–478–14596–0 (web PDF)
This is a client report commissioned by Nelson/Marlborough Conservancy and funded from the
Science Advice Fund. It was prepared for publication by the Publishing Team; editing and layout by
Lynette Clelland. Publication was approved by the General Manager, Research and Development Group,
Department of Conservation, Wellington, New Zealand.
In the interest of forest conservation, we support paperless electronic publishing. When printing,
recycled paper is used wherever possible.
C ontents
Abstract
5
1.
Introduction
6
2.
What is known about dwarf galaxias?
6
3.
Methods
9
4.
Results
10
4.1
Physical habitat
10
4.2
Observed fish behaviour
12
4.3
Dwarf galaxias population
12
5.
Discussion
13
5.1
16
Link with black-fronted terns
6.
Conclusions
18
7.
Acknowledgements
19
8.
References
19
Effects of low flow on
dwarf galaxias and their
habitat in the Wairau River
Joe Hay
Cawthron Institute, 98 Halifax Street East, Private Bag 2, Nelson Mail Centre,
Nelson 7042, New Zealand
Email: [email protected]
A bstract
Trustpower’s proposed power scheme on the braided Wairau River, Marlborough,
New Zealand, would divert up to 40 m 3/s of water through a 50-km canal, leaving
a residual flow of 10–20 m 3/s in the affected reaches of the river. Two-dimensional
flow modelling has predicted that at flows of 15–20 m 3/s, the flow in the affected
reaches would be largely confined to the main channel, effectively eliminating
shallow side-braid habitat, with an associated decline in available habitat for
dwarf galaxias (Galaxias divergens), a small endemic fish that inhabits these sidebraids. This report investigates changes in a population of dwarf galaxias as flow
declined over one summer season, in habitats provided by a seep-fed side-braid
of the Wairau River. The study confirmed that as flow dropped to approximately
20 m3/s, habitat availability was reduced, and the braid became isolated from
the main channel. This occurred alongside an apparent natural reduction in the
abundance of dwarf galaxias, from a post-spawning abundance of predominantly
juvenile fish. The ecology of dwarf galaxias appears to enable them to tolerate
natural low-flow events. However, had this flow reduction occurred earlier in the
year (or been sustained year round), it may have had a negative impact on the
dwarf galaxias population. Side-braid habitat is likely to be vital for the survival of
dwarf galaxias in the Wairau River, providing habitat, abundant invertebrate food
and refuge from both physical and biological disturbances. Side-braids also appear
to provide ideal foraging habitat for black-fronted terns (Sterna albostriata)
which feed on dwarf galaxias and other small fishes during their breeding season.
Ensuring that water abstraction at the proposed water intake does not cause the
frequency and duration of occurrence of flows ≤ 22 m3/s to exceed natural rates
should avoid exacerbating low-flow impacts on dwarf galaxias and foraging area
for black-fronted terns downstream.
Keywords: dwarf galaxias, Galaxias divergens, non-migratory galaxiids, habitat,
flow, low flow, flow reduction, braided river, black-fronted tern, Sterna albostriata,
Wairau River
© April 2009, New Zealand Department of Conservation. This paper may be cited as:
Hay, J. 2009: Effects of low flow on dwarf galaxias and their habitat in the Wairau River. DOC
Research & Development Series 309. Department of Conservation, Wellington. 20 p.
DOC Research & Development Series 309
5
1. Introduction
This report investigates changes in a population of dwarf galaxias (Galaxias
divergens) in a variety of habitats provided by a small, seep-fed side braid of
the Wairau River, Marlborough (Fig. 1), as flow declined over the summer. The
study aimed to increase the level of understanding of the potential responses of
G. divergens, and their habitat, to reductions in river flow.
The need for more information about the flow requirements of G. divergens was
prompted by Trustpower’s proposal to develop a hydroelectric power scheme
in the Wairau Valley. This hydro scheme would divert some of the Wairau River
flow (up to 40 m 3/s; Trustpower 2005) into a canal which would feed through six
new power stations before discharging back into the Wairau River approximately
50 km downstream. Trustpower is proposing seasonal minimum flows in the
range 10–20 m 3/s for the affected section of river (Ryder & Keesing 2005).
2. What is known about
dwarf galaxias?
Dwarf galaxias (Galaxias divergens) is a species of small freshwater fish, found
in parts of the North and South Islands of New Zealand. It was first described
from the Maruia River in the northern South Island by Stokell (1959). Dwarf
galaxias have cryptic habits, making them difficult to study. Partly as a result of
this, their ecology is poorly understood.
The dwarf galaxias is a member of the family Galaxiidae, best known in
New Zealand for its sea-migratory (diadromous) members whose juveniles
comprise whitebait runs. Dwarf galaxias is one of several non-diadromous
members of the family (i.e. individuals complete their entire life-cycle in
freshwater), so they do not contribute to the whitebait fishery. These fish
are known as dwarf galaxias because of their diminutive size—they generally
grow to only about 70 mm long (maximum length recorded is 87 mm; with fish
from the north of the South Island tending to be larger than those elsewhere in
New Zealand (McDowall 2000)). This compares with average lengths among
closely related species of 80–85 mm (alpine galaxias, G. paucispondylus) and
100–120 mm (Canterbury galaxias, G. vulgaris) (McDowall 2000).
The scientific name—Galaxias divergens—alludes to morphological differences
between this species and the majority of other members of the Galaxiidae
(G. divergens has six rather than seven pelvic fin rays and 15 caudal fin rays;
Stokell 1959; NIWA 2005).
Dwarf galaxias closely resembles the Canterbury galaxias (G. vulgaris), and
the ranges of the two species overlap in Marlborough (McDowall 2000). The
range of the dwarf galaxias extends from the Hokitika River catchment on the
West Coast (Fig. 2) through Nelson and Marlborough, Wellington and Manawatu,
and into the southern Hawke’s Bay, with outlying populations in the Hauraki
Plains and Bay of Plenty.
6
Hay—Dwarf galaxias habitat requirements
Recent genetic analyses have identified five genetically distinct populations
within this range (Allibone 2000, 2002). The Wairau River fish belong to the
most widespread of these populations, extending from southern Marlborough
north to as far as Hawkes Bay, suggesting a land link between North Island and
South Island rivers at some time in the past, probably when sea levels were low
during the last glaciation.
Dwarf glaxias can be found living among cobble and gravel substrates in riffles of
small streams, and in the shallow riffle margins of larger rivers, mainly in foothill
catchments (McDowall 2000).
Individuals are thought to live 2–3 years (McDowall 2000), with female fish
apparently outliving male fish. As a result, females tend to be numerically
dominant in the second- and third-year age classes (Hopkins 1971). Dwarf
galaxias generally mature at 2 years old, and spawn during spring (and possibly
autumn) (McDowall 2000). In a study in Hinau Stream, Wairarapa (southern
North Island), ripe fish were found from September to December, but not by
mid-January (Hopkins 1971). Although their spawning sites have not yet been
discovered, it is thought likely that they are in the same area as adult habitat
(McDowall 2000). Approximately 100–250 eggs, each about 2 mm in diameter,
are laid. Observations of eggs laid by captive fish in aquaria suggest that dwarf
galaxias may lay their eggs singly, or in small groups attached to stones (Hopkins
1971) (cf. Canterbury galaxias, G. vulgaris, which lays its eggs in large masses on
big stones; McDowall 2000), although this supposition is based on limited data.
Figure 1. Map of Wairau
River showing study reach,
flow recorder sites and the
approximate location of the
proposed power scheme.
Tuamarina flow recorder
N
N
te
ul
rR
iver
Study site
out
te c
ma
xi
pro
lr
ana
Ap
Blenheim
e
ive
p ai R
iho
a
W
r
Go
Proposed outlet site
Proposed intake site
Bran
c
Wairau R
ive
r
hR
iv
er
Wash Bridge
SH1
Dip Flat flow recorder
10 km
DOC Research & Development Series 309
7
Figure 2. Dwarf galaxias
(Galaxias divergens)
distribution map, based
on records from the
New Zealand Freshwater
Fisheries Database
(administered by the
National Institute of Water
and Atmospheric Research;
accessed on 21 Jan 2009).
N
Bay of Plenty
Hawkes Bay
Manawatu
Nelson
Wellington
Marlborough
West Coast
200 km
Small shoals of larvae and juveniles may be found around stream margins during
spring and summer (McDowall 2000). Hopkins (1971) found fry from late
September until early March in Hinau Stream, but they were most abundant in
November and December. Interestingly, Hopkins always found fry earlier in a
small, isolated, spring-fed pool than in the main channel of the stream. Fry were
seen in the spring-fed pool in September, but were not evident in the main
channel until October. Hopkins also noted that fry were able to swim immediately
upon hatching, and probably shoal straight after hatching as well.
Juveniles tend to hold station (i.e. maintain their location relative to the ground)
in the water column during their early life, but become more benthic and cryptic
at between 25 mm and 35 mm long (aged 2–3 months) (Hopkins 1971).
8
Hay—Dwarf galaxias habitat requirements
The diet of dwarf galaxias is mainly benthic invertebrates (especially chironomid
fly larvae and Deleatidium mayfly nymphs) (McDowall 2000). However, terrestrial
prey recorded in stomach contents suggests that the fish feed from the drift as
well as the benthos (Hopkins 1971). Hopkins (1971) recorded a dietary shift
from mainly chironomids eaten by small juveniles, to mainly Deleatidium as
fish grew larger. He also noted that the stomachs were fullest in the afternoon,
suggesting that dwarf galaxias are predominantly diurnal feeders.
In common with some other members of the Galaxiidae, adult dwarf galaxias
are able to burrow into gravel, presumably to seek refuge from predators and
extreme flow conditions. Under experimental conditions, dwarf galaxias, with
their smaller body form, burrowed faster and deeper than larger alpine galaxias
(G. paucispondylus), Canterbury galaxias, and upland bullies (Gobiomorphus
breviceps) in response to reduced water levels (Hartman 1990). This adaptation
presumably allows them to take refuge amongst moist gravels during relatively
short-term drought events, as well as avoiding washout during floods.
Along with the majority of non-migratory Galaxiid species in New Zealand, dwarf
galaxias is considered to be nationally threatened and is listed by the Department
of Conservation (DOC) as being in ‘Gradual Decline’ (DOC 2004). A variety of
causes have been implicated in the decline of the species, including landuse
change, impacts of introduced species and water abstraction.
3. Methods
On 14 January 2005, nine reaches (15–58 m in length) in a seep-fed side-braid
of the Wairau River (immediately downstream of its confluence with the Branch
River; Fig. 1) were surveyed. The reaches were selected to cover the range of
habitats that dwarf galaxias were observed to utilise in this braid. Three transects
were located within each reach (at the top, middle and bottom of the reach).
These were marked with ‘dazzle’ spray-paint and recorded using GPS, to facilitate
their location on subsequent visits.
Flow in the Wairau River during this initial survey was approximately 75 m 3/s at
Tuamarina and 26 m 3/s at Dip Flat (Fig. 1). The flow experienced at our sampling
location would have been somewhere between these two recorded flows, but
closer to the Dip Flat flow because of its proximity and the location of major
tributary inflows (Fig. 1).
The survey involved placing a 0.25-m 2 quadrat approximately mid-channel at
each transect. An observer then took up a position within viewing distance of
each quadrat (i.e. one observer per quadrat) and waited for 3 minutes for fish to
resume their normal activity. The area of river delineated by the quadrat was then
observed for 5 minutes, and the maximum number of fish, of any given species,
that was observed within the quadrat area, at any time during the monitoring
period, was noted. This approach was adopted to avoid double counting of these
sometimes highly-mobile fish as they moved in and out of the quadrat during
the observation period. The resulting fish counts provide a relative index of
abundance that can be used to make comparisons over time, rather than providing
an estimate of actual fish abundance or density.
DOC Research & Development Series 309
9
Observers also estimated the size of the fish they counted using a known scale to
calibrate their estimates. This allowed the size-class composition of the counted
fish to be estimated.
Physical habitat observations were also made for each reach. These included
riparian cover, substrate composition, wetted width at each transect, reach
length, water temperature at the top and bottom of each reach, maximum depth
within each quadrat and water velocity at that point.
Surveys were repeated on 25 January, 1 March and 11 April 2005. River flows
on these follow-up visits were approximately 30 m 3/s, 20 m3/s, and 35 m 3/s
respectively at Tuamarina, and 16 m 3/s, 13 m3/s, and 20 m 3/s respectively at Dip
Flat. It was intended to conduct the final survey at a lower flow. The recorded
flow at Tuamarina available on the day of the last survey was 11 m 3/s, but the
flow recorder was subsequently found to be faulty. Reliable flow recorder
measurements for this site were subsequently obtained from another recorder
at the site.
Water clarity was good during all of the surveys, although ripples on the water
surface caused by wind gusts periodically disrupted visibility, particularly in the
afternoons.
4. Results
4.1
P h y sical habitat
Between the first survey (14 January) and second survey (25 January), water
levels had noticeably reduced only at the lower two reaches, those closest to the
outlet into the main channel. There was a noticeable increase in the abundance
of filamentous green and blue-green algae in most reaches.
By the third visit (1 March), the water level was noticeably reduced throughout
the survey area, with two reaches (and one transect in another) totally dry
(Fig. 3). One of these reaches was still dry during the final survey (11 April; Fig. 4,
Table 1), despite the increase in flow recorded (Table 1). Drying had effectively
isolated the whole side-braid from the main channel.
There was also a noticeable decrease in the abundance of filamentous algae in
most reaches by the 1 March survey, and an increase in silt.
Between the third and fourth surveys, the survey area experienced a large flood
(peaking at approximately 1100 m 3/s at Tuamarina). This changed the riverbed
morphology (channel width, depth, substrate composition) to varying degrees
at most of the survey reaches. It also removed filamentous algae and macrophyte
cover from the majority of reaches. However, one survey reach appeared to have
been largely untouched by the flood.
There was a marked reduction in the wetted area of most survey reaches during the
first three surveys (Table 1), corresponding with reducing flows over this period.
As flow dropped between the second and third surveys (from 16 m 3/s to 13 m3/s at
Dip Flat), the total wetted area in the surveyed reaches declined by approximately
28%. However, the changes in bed form associated with the large flood (between
10
Hay—Dwarf galaxias habitat requirements
Figure 3. The top survey
reach as flow decreased
over the four survey visits,
14 Jan 2005 (top left),
25 Jan 2005 (top right),
1 March 2005 (bottom left)
and 11 April 2005 (bottom
right).
Figure 4. The bottom
survey reach which
was dewatered as flow
decreased during the
summer, 25 Jan 2005 (left)
and 11 April 2005 (right).
T able 1 . C hanges in the area of available wetted habitat in the surve y ed reaches of a seep - fed
side braid of the W airau R iver , as flow declined through the 2 0 0 5 summer season .
Wetted area (m 2 )
Survey
date
Flow (m 3 /s)in each Reach (numbered)
(2005)tuamarinadip flat
1
2
3
4 & 5
6
7
8
9
Total
14 Jan
76
26
235.9
67.1
183.3
137.1
71.2
280.0
118.8
63.8
1157.2
25 Jan
30
16
248.8
70.2
183.3
134.0
71.2
246.4
103.5
48.1
1105.5
1 Mar
20
13
164.3
0.0
165.9
131.2
65.0
232.8
0.0
38.2
797.4
11 Apr
36
20
115.0
152.5
148.5
141.7
66.5
308.0
0.0
53.4
985.5
the third and fourth surveys) produced some broad shallow areas in reaches that
had formerly been relatively narrow. This meant that the wetted areas recorded
during this last survey were not directly comparable with those recorded on earlier
visits. Similar decreasing trends were also evident in the depth and velocity records
from the surveyed reaches, over the first three visits.
DOC Research & Development Series 309
11
4.2
O bserved fish behaviour
Individual galaxiids, when disturbed, tended to dart around before taking cover;
whereas several individuals, when disturbed at the same time, would often form
shoals which would generally disperse after a few minutes. The individuals
would then often hold station in the water column in much the same fashion as
drift-feeding trout.
4.3
D warf gala x ias population
The numbers of dwarf galaxias counted during the surveys declined over the
summer (Fig. 5), although there was also a large degree of spatial variability in
counts. Casual observations of fish abundance made by observers as they moved
between reaches also indicated that overall abundance was declining. A decline
in numbers is common in juvenile fish in their first year of life, due to densitydependent mortality and emigration (Elliot 1994; Begon et al. 1996).
The growth of the new season’s cohort as the summer progressed was evident in
the data (Fig. 5). Counts early in the season included a range of different-sized fish,
from fry (< 15 mm in length) to adults, and with juveniles in the size range 26 mm
and 40 mm particularly prevalent. However, by the last survey adults (> 50 mm)
predominated and the fish smaller than 30 mm were no longer present. Hence,
counts early in the season comprised large numbers of predominantly small fish,
while counts in the later part of the season comprised smaller numbers of larger
fish.
By the fourth visit (11 April), fish counts (as well as the numbers of fish seen
shoaling by observers as they moved between sampling sites) were appreciably
lower than on previous trips—in fact, fish were absent from all but a couple of
isolated locations by the time of the last survey). This was despite river flows on
the fourth visit being greater than they had been during the third visit. All but one
of the observed fish in the last visit were of adult size (i.e. ≥ 60 mm long) (Fig. 5).
80
Count (total
(total accross
across all
Count
all survey
survey reaches)
reaches)
Figure 5. Changes in
observed dwarf galaxias
(Galaxias divergens)
abundance and population
structure during summer
2005.
14 Jan 2005
70
25 Jan 2005
60
01 Mar 2005
11 Apr 2005
50
40
30
20
10
0
Total
< 15
16–20
21–25
26–30
31–35
Length class (mm)
12
Hay—Dwarf galaxias habitat requirements
36–40
41–50
> 51
As the summer progressed and sections of the surveyed area dried out, large
shoals of dwarf galaxias were occasionally observed in pools upstream of the
dry reaches. Dunn (2003) noted a similar occurrence, with alpine galaxias and
Canterbury galaxias seen to concentrate in pools as a reach dried out, leading to
apparent competition for space.
One adult dwarf galaxias was found floating dead below one of the reaches on
the third survey visit (1 March). There was no evidence of injury and the cause
of death was not apparent.
5. Discussion
This study showed that as flow at Dip Flat declined from 26 m 3/s to 13 m 3/s,
the area of wetted habitat available to dwarf galaxias in a seep-fed side-braid
of the Wairau River was substantially reduced, and the remaining wetted area
became isolated from the main channel. In this case the flow reduction occurred
naturally, as a result of low summer rainfall, and coincided with an expected
natural decline in number of dwarf galaxias over their first summer of life (i.e.
from an initial high post-spawning abundance of fry). In order to have persisted
in this system, this species must be well adapted to frequently occurring lowand high-flow events. The flow range recorded at Dip Flat over this period was
above the mean annual low flow for this site (approximately 8.6 m 3/s; Mitchell
2004), so flows as low as that recorded could be expected to occur in most years.
It is unlikely, therefore, that the reduced flow recorded during this study had
any significant long-term negative effect on the dwarf galaxias population in the
Wairau River.
However, artificially reduced flows earlier in the season might be expected to
have more of an impact. Extremes of flow (i.e. floods and droughts) during critical
spawning and fry-rearing periods have been found to limit recruitment of nonmigratory galaxiids, strongly influencing population size structure (Dunn 2003).
Dunn (2003) observed high mortality among alpine galaxias and Canterbury galaxias
in Dry Stream, Castle Hill, Canterbury, New Zealand, during low-flow events in
spring. The effects of low flow may have compounded the already high stress levels
associated with reproduction, resulting in increased mortality (Dunn 2003).
In common with other non-migratory galaxiids, dwarf galaxias appear to have
physiological and behavioural adaptations that allow them to survive shortduration drying events. As mentioned above, dwarf galaxias are more proficient
burrowers than their close relatives, alpine and Canterbury galaxias (Hartman
1990). The latter species have the ability to survive out of water for a week or
more, if kept under moist conditions (Dunn 2003). It seems reasonable to expect
that dwarf galaxias should have a similar ability. These sorts of adaptations would
be necessary for the long-term survival of any fish species faced with the highly
variable flows that many of New Zealand’s hill country streams experience.
However, these behavioural and physiological adaptations are only likely to
be effective against relatively short-term low-flow events, and would probably
not allow dwarf galaxias to endure chronic low flows caused by artificial flow
DOC Research & Development Series 309
13
diversions. The only viable survival option the fish would have under chronic
low-flow conditions would be to move to more suitable habitat elsewhere, if it
is available. In a braided river, emigration to deeper, permanent channels might
increase predation risk from fish predators such as trout and eels.
The habitat provided by side braids may be of particularly high value for small
fish living in large rivers such as the Wairau. Isolated side braids are less likely to
experience the same frequency of floods as the main channel, and may therefore
provide suitable flow refugia for fry. This effect was evident in our study following
the large flood event of 25–27 March 2005. Although the majority of the surveyed
reaches appeared to have been scoured to some degree and morphological changes
were evident in places, one of the survey reaches was apparently unaffected. This
reach still retained reasonably substantial filamentous algal growths, as well as
reasonable numbers of dwarf galaxias, suggesting that it had acted as an effective
flood refuge, even during this large flood. As a consequence of their stability, seepfed side braids also generally have higher densities of the benthic invertebrates,
which are the main food of dwarf galaxias (Hughey et al. 1989).
The refuge from high water velocities (during floods) provided by side braids
is likely to be of particular value during the spawning season. In a study on
factors affecting recruitment of alpine and Canterbury galaxias, Jellyman (2004)
found that the amount of slow, backwater habitat available was one of the most
important factors determining initial fry density. Young galaxiids have low
velocity tolerance (10-mm-long fry are able to hold station against a velocity of
only 0.1 m/s; Jellyman 2004), and therefore require the velocity refuge offered
by slow backwater habitat to prevent them from being washed downstream. This
corresponds well with observations made during the present study, where shoals
of small fry were generally seen in lower-velocity water, out of the main current.
Furthermore, Jellyman (2004) found that washout during flooding was a major
cause of decline in fry density.
Side braids may also provide a refuge for dwarf galaxias from predatory fish,
particularly trout. The shallow depths and seep-fed nature of many side braids
may limit their accessibility to large trout, which are likely to be piscivorous
(brown trout tend to become piscivorous when they grow > 300 mm long; Keeley
& Grant 2001). The physical conditions experienced in side braids may also
exclude trout (Dunn 2003). Water temperatures recorded during our surveys
were as high as 27.7°C, well into the lethal range for brown trout (Elliot 1994).
During the January surveys, the water temperature was almost universally > 19°C
(brown trout are known to stop feeding at temperatures above 19°C; Elliot 1994),
except for the top reaches, where all flow was seepage from the substrate and
temperatures ranged between 14°C and 18°C.
A recent biological survey of the Wairau River, undertaken for Trustpower, found
that ‘…secondary and tertiary braids and seepages in sampled reaches of the Wairau
have very high native fishery values, strong macroinvertebrate communities, good
quality water and healthy periphyton communities’ (Ryder & Keesing 2005: 109).
The authors suggested that these side-braid habitats are very important for galaxiids
and bullies, along with other native fish species. In an earlier version of their report
it was speculated that if habitat availability in these side-braids was reduced by
lower flows, the remaining habitat would probably not be capable of maintaining
the same biomass (Boffa Miskell & Ryder 2004).
14
Hay—Dwarf galaxias habitat requirements
Two-dimensional flow modelling (undertaken as part of the feasibility study for
Trustpower’s proposed hydro scheme) predicts that flowing water would be
restricted to the main channel in the reaches affected by the abstraction when
the overall river flow drops to between 15 and 20 m 3/s. Flows in this range would
largely eliminate side-braid habitat at the upstream end of the reaches affected by
abstraction. Further downstream, as input from tributaries increases the residual
flow in the river, some side-braids are predicted to remain wetted. However,
this remnant side-braid habitat is predicted to be largely isolated from the main
channel over this flow range. Connectivity between side-braid habitat and the
main channel is likely to be critical to the dispersal of juvenile dwarf galaxias,
allowing them to colonise new areas of suitable habitat as these are created in
the braided river system. The modelling predicted that reducing flow to these
levels would result in a pronounced drop in habitat availability for dwarf galaxias
(Ryder & Keesing 2005).
Our observations indicated that the side-braid was indeed beginning to be
dewatered at flows in the order of 15–20 m 3/s, and that connectivity with the
main channel was lost before flows reached these levels (i.e. at higher flows).
In a report on the hydrology of the Wairau catchment, prepared as part of
Trustpower’s feasibility study, Mitchell (2004) presented the following regression
for predicting flow (Q) at Wash Bridge (see Fig. 1 for location) from flow at Dip
Flat:
QWash Bridge (m3/s) = 1.351 Q Dip Flat (m3/s) – 0.6497 (r2 = 0.9505).
Mitchell used this relationship to predict flow at the proposed intake to the
hydropower scheme by adding 0.3–0.6 m 3/s to account for a reasonably constant
inflow between Wash Bridge and the proposed intake. Based on this method, the
flow range 13–16 m 3/s at Dip Flat, over which partial dewatering of the side-braid
was observed, causing the side-braid to become isolated from the main channel,
would be equivalent to flows of approximately 17–22 m 3/s at the proposed hydro
scheme canal intake.
This flow range (i.e. 17–22 m 3/s at the proposed intake) is substantially higher
than the natural mean annual low flow (MALF) in the section of the Wairau River
affected by the proposed hydro scheme (according to Hudson et al. (2004), the
MALF in the reaches affected by the proposed scheme ranges between 10 m 3/s
and 15 m 3/s, depending on the distance downstream 1). Therefore, dwarf galaxias
populations living in side-braids of the Wairau River must be adapted to drying
events of this magnitude. However, while these populations of dwarf galaxias
must be resilient to low-flow events of relatively short duration, this resilience
may not be sufficient to enable them to survive such low levels over prolonged
periods, particularly if flow reductions occur during the critical spawning and
dispersal life stages. Further research is needed to ascertain the magnitude and
duration of low-flow events that dwarf galaxias populations are able to tolerate.
However, in the meantime, so long as flows ≤ 22 m3/s at the proposed water
intake (equivalent to approximately 16 m 3/s at Dip Flat) do not exceed their
natural frequency and duration of occurrence, then low-flow impacts on dwarf
galaxias habitat and populations should not be exacerbated.
1
Flow accrues gradually as a river progresses downstream, due to natural inflows from streams
and the substrate. Therefore, the flow at the top of a particular reach will normally be less than
that at the bottom of the reach. As a result, mean annual low flow also increases downstream
through the affected reach.
DOC Research & Development Series 309
15
5.1
L ink with black - fronted terns
The dwarf galaxias population in the Wairau River may be an important food
source for birds. In particular, the aggregations of juvenile dwarf galaxias that
occur in the river shallows during early summer may contribute to the breeding
success of black-fronted terns (Sterna albostriata) nesting in the area. Food
availability may be important in the timing of breeding seasons and in population
regulation (through chick survival) in migratory birds (Hughey 1985).
The black-fronted tern is classified as ‘Nationally Endangered’ (Hitchmough
et al. 2007). It breeds only on the braided rivers of the South Island, with the
Wairau River being home to the largest known population (Keedwell 2002). The
entire black-fronted tern population is estimated to number less than 10 000
birds (Keedwell 2002), with the Wairau River supporting a population of > 1000
of these. In the 2004/05 breeding season, the Wairau breeding population
included at least four colonies in the area likely to be affected by reduced flows
if Trustpower’s proposed hydro scheme goes ahead (Bell 2005).
Although adult black-fronted terns feed mainly on aquatic insects, fish are an
important food source for their chicks, comprising the bulk of their diet in terms
of energy intake (Lalas 1977). The high energy content of fish makes them an
attractive food source for chicks. This is particularly true for black-fronted terns,
since they feed their young whole prey, rather than regurgitated prey, so the
energy intake of chicks is limited by the time it takes for the parents to deliver
prey items to the nest (Lalas 1977). Lalas calculated the maximum energy intake
of chicks fed entirely on insects at 3–9 kJ/h. This equates to approximately 60–80
insects (i.e. 60–80 return trips to the nest for the adult bird), but is equivalent to
only 0.7–2 fish. Thus, feeding chicks on fish reduces the number of trips to the
nest an adult has to make, therefore minimising their energy expenditure.
Black-fronted terns are capable of handling only relatively small fish. The largest
fish that Lalas (1977) saw successfully captured by a black-fronted tern was
80 mm long. Terns are also limited in the depth to which they can dive for
prey; most fish are taken over a depth range 0.05–0.2 m, but some are taken
up to a maximum of about 0.4 m (Lalas 1977). These criteria suggest dwarf
galaxias should be an ideal prey species and shallow side-braids an ideal foraging
habitat for black-fronted terns. Indeed, Lalas (1977) observed black-fronted terns
catching galaxiid species (Longjaw galaxias (G. prognathus) and Canterbury
galaxias), along with trout fry and upland bullies, in the Waitaki Catchment. He
also noted that terns foraged for fish mainly along minor, slower-flowing sidebraids, especially during September to November.
Black-fronted terns on the Wairau River are often seen hunting along the shallow
margins of the river, the habitat best suited to dwarf galaxias. There have also
been anecdotal reports of terns carrying small fish resembling dwarf galaxias,
and these sightings coincided with the period that terns would be feeding chicks
(J. Clayton-Greene & S. Cranwell, DOC, pers. comm.).
Breeding and chick-rearing by black-fronted terns coincides with the period
when densities of juvenile dwarf galaxias are at their peak (Fig. 6). It appears to
be relatively common for terns to breed near, and at times which coincide with,
a rich food source provided by aggregations of juvenile fish. Taylor & Roe (2004)
found that little terns (Sterna albifrons sinensis) breeding on Rigby Island,
Gippsland Lakes, in south-east Australia, fed their chicks entirely on juvenile fish
(with prey length ranging between 7.5 mm and 82.5 mm). Furthermore, breeding
16
Hay—Dwarf galaxias habitat requirements
Figure 6. Comparison
between the timing of
dwarf galaxias (Galaxias
divergens) and blackfronted tern (Sterna
albostriata) life history
stages.
success among these birds was significantly reduced in the season following
mass mortality of juvenile pilchards, one of their most important prey species.
Nesting colonies of Caspian terns on the Columbia River, USA, were found to
feed almost exclusively on salmonid smolts (Collis et al. 2002). These juvenile
fish accounted for 73–81% of identifiable prey items in local Caspian tern diets,
and were the primary prey item for entire tern breeding colonies, especially
during pre-laying, laying and incubation (which coincided with the peak in
seaward migration of smolts).
Variability in the availability of food has been associated with variations in both
numbers of terns breeding and breeding success (Taylor & Roe 2004). Taylor &
Roe (2004) postulated that short supply of food, and consequent longer periods
away from the nest foraging, may lead to increased predation risk for nestlings
and, ultimately, that food shortages might lead to the abandonment of colonies.
In a study of black-fronted terns nesting over four years in the Ohau River,
south Canterbury, Keedwell (2002) found at least 4.1% of chick mortality was
attributable to starvation. Also, nest desertion during incubation accounted for
21.4% of nest failures in this study (Keedwell 2002). Nest desertion was second
only to predation as a cause of nest failure (24.6% of failures).
Substantial decline in the availability of dwarf galaxias (as well as other small,
shallow-water-dwelling fish species, e.g. upland bullies) in the proximity of
black-fronted tern colonies in the Wairau River would risk reducing the breeding
success of black-fronted terns. Flow reduction has the potential to do this through
reduced habitat availability for dwarf galaxias. If flows were reduced to a level that
confined flow to the main channel, eliminating side-braids, black-fronted terns
would be confronted with a reduction in suitable foraging habitat. This would be
compounded if the abundance of their prey (dwarf galaxias) was also reduced, as
would be expected if the habitat provided by side-braids was eliminated. There
is also evidence that nest predation pressure is greater in areas connected to the
mainland than on islands in braided rivers (Pascoe 1995; Rebergen et al. 1998).
This suggests that reduced flow could also have a direct impact on black-fronted
terns, by removing the nesting refuges safe from ground-based predators that
islands in active braided river systems provide. There is some limited evidence
that nest predation rates on the Wairau River may be higher when levels of flow
around islands are low (Bell 2005). However, further investigation is required to
assess the veracity of this potential relationship.
DOC Research & Development Series 309
17
6. Conclusions
Trustpower’s proposal to abstract water from the Wairau River for a hydropower
scheme would result in river flows being reduced significantly over approximately
50 km of the river. Two-dimensional flow modelling has predicted that if flow
is reduced below 15–20 m3/s, the residual flow in the river will be confined to
the main channel, effectively eliminating shallow side-braid habitat, or isolating
it from the main channel, with an associated decline in available dwarf galaxias
habitat.
The present study tracked changes in the physical habitat and relative abundance
of dwarf galaxias as flow declined over summer 2005 in a seep-fed side-braid.
It confirmed that as flow dropped to approximately 20 m3/s at the study reach, the
area of available habitat in the side-braid was reduced (declining by approximately
28% as flows recorded at Dip Flat dropped from 16 m3/s to 13 m3/s, equivalent
to approximately 22 m3/s to 17 m3/s at the proposed hydro scheme intake).
This reduction in habitat area occurred alongside an expected natural reduction
in the abundance of dwarf galaxias, from a high post-spawning abundance of
predominantly juvenile fish. However, if this reduction in flow occurred earlier
in the year (or was sustained year round), it would be likely to have a negative
impact on the dwarf galaxias populations inhabiting side-braids.
As well as providing appropriate physical habitat and abundant invertebrate food
species for dwarf galaxias, side-braid habitats likely provide a refuge for riverresident galaxiids from both physical (e.g. droughts and floods) and biological
(e.g. predatory fish) disturbances. They may be of particular importance as a
water velocity refuge for juvenile fish, which have limited swimming abilities.
Side-braids also appear to provide ideal foraging habitat for black-fronted terns
during their breeding season.
Further research is required to assess what level of habitat reduction might be
tolerated by the dwarf galaxias populations in the Wairau River, and how habitat
reduction might affect the productivity of these populations. More information
is also required on the magnitude and duration of the low-flow events that these
populations are capable of enduring. Finally, further investigation is needed into
the role played by dwarf galaxias, and other small fishes found in shallow sidebraid habitats, in the breeding success of the threatened black-fronted tern.
If flow reductions decrease the availability of edgewater habitat provided by
shallow side-braids in the Wairau River, this would not only impact on dwarf
galaxias, but also potentially affect the threatened black-fronted tern population.
Assuming that other side-braids in the reaches affected by water abstraction
for the proposed hydropower scheme have similar responses to low flows,
our observations indicate that maintaining the natural frequency and duration
of occurrence of flows ≤ 22 m3/s at the proposed water intake (equivalent to
about 16 m3/s at Dip Flat) should avoid exacerbating low-flow impacts on dwarf
galaxias habitat and populations, as well as foraging area for black-fronted terns
feeding on dwarf galaxias, in these reaches.
18
Hay—Dwarf galaxias habitat requirements
7. Acknowledgements
Rowan Strickland, Jan Clayton-Greene, Claire Wooldridge-Way, Tom Stein and
Warren Mitchell provided assistance with fieldwork, and John Hayes reviewed
the draft report. Run-holders Judith and Ken Reid provided access over their
property. Steve Cranwell, DOC, provided information on black-fronted terns
and NIWA provided the confirmed flow data for the Wairau River at Tuamarina
(Barnetts Bank) and Dip Flat.
8. References
Allibone, R.M. 2000: Population structure of Galaxias divergens, a widespread non-migratory
galaxias in New Zealand. Paper presented at Fresh perspectives: joint conference of
the New Zealand Limnological Society, the New Zealand Hydrological Society and the
Meteorological Society of New Zealand, November 2000, Christchurch, New Zealand.
www.niwa.cri.nz/rc/prog/freshbiodiversity/info/allibone
Allibone, R.M. 2002: Dealing with diversity dwarf galaxias style. Water & Atmosphere 10: 18–19.
Begon, M.; Harper, J.L.; Townsend, C.R. 1996: Ecology: individuals, populations and communities
(third edition). Blackwell Science, Oxford, UK. 1068 p.
Boffa Miskell Ltd.; Ryder Consulting Ltd 2004: Proposed Wairau Hydroelectric Power Scheme: ecological
investigations of the Wairau River main stem (final draft). Unpublished report. 73 p.
Bell, M. 2005: Breeding success of black-fronted tern on the Wairau River. Unpublished report
prepared for the South Marlborough Area Office, Nelson/Marlborough Conservancy,
Department of Conservation. February 2005. 10 p.
Collis, K.; Roby, D.D.; Craig, D.P.; Adamany, S.; Adkins, J.Y.; Lyons, D.E. 2002: Colony size and diet
composition of piscivorous waterbirds on the lower Columbia River: implications for losses
of juvenile salmonids to avian predation. Transactions of the American Fisheries Society
131: 537–550.
Elliott, J.M. 1994: Quantitative ecology and the brown trout. Oxford University Press, Oxford, UK.
286 p.
DOC (Department of Conservation) 2004: New Zealand non-migratory galaxiid fishes recovery plan 2003–
13. Threatened Species Recovery Plan 53. Department of Conservation, Wellington. 45 p.
Dunn, N.R. 2003: The effects of extremes in flow on alpine (Galaxias paucispondylus) and Canterbury
(G. vulgaris) galaxias. Unpublished MSc thesis, University of Canterbury, Christchurch. 174 p.
Hartman, G. 1990: Ability of some New Zealand fishes to burrow in gravel. Freshwater Catch 44:
15–16.
Hitchmough, R.; Bull, L.; Cromarty, P. (Comps) 2007: New Zealand Threat Classification System lists
2005. Department of Conservation, Wellington, New Zealand. 194 p.
Hopkins, C.L. 1971: Life history of Galaxias divergens (Salmonoidea: Galaxiidae). New Zealand
Journal of Marine and Freshwater Research 5: 41–57.
Hudson, H.R.; McMillan, D.A; Grey, A. 2004: Hydraulic and habitat simulation, Wairau River,
Marlborough (final draft). EMA 2004–02, Environmental Management Associates,
Christchurch. 59 p.
DOC Research & Development Series 309
19
Hughey, K.F.D. 1985: Hydrological factors influencing the ecology of riverbed breeding birds on
the plains’ reaches of Canterbury’s braided rivers. Unpublished PhD thesis, University of
Canterbury, Chistchurch, New Zealand. 255 p.
Hughey, K.; Fraser, B.R.; Hudson, L.G. 1989: Aquatic invertebrates in two Canterbury rivers related to
bird feeding and water development impacts. Science & Research Series No. 12. Department
of Conservation, Wellington. 61 p.
Jellyman, P.G. 2004: Fry survival of alpine (Galaxias paucispondylus) and Canterbury (G. vulgaris)
galaxids. Unpublished BSc Hons thesis, University of Canterbury, Christchurch, New Zealand.
68 p.
Keedwell, R.J. 2002: Black-fronted terns and banded dotterels: causes of mortality and comparisons
of survival. Unpublished PhD thesis, Massey University, Palmerston North, New Zealand.
194 p.
Keeley, E.R.; Grant, J.W.A. 2001: Prey size of salmonid fishes in streams, lakes, and oceans. Canadian
Journal of Fisheries and Aquatic Sciences 58: 1122–1132.
Lalas, C. 1977: Food and feeding behaviour of the black-fronted tern, Chlidonias hybrida albostriatus.
Unpublished MSc thesis, University of Otago, Dunedin, New Zealand. 164 p.
McDowall, R.M. 2000: The Reed guide to New Zealand freshwater fishes. Reed, Auckland,
New Zealand. 224 p.
Mitchell, P. 2004: Wairau Valley hydo-electric power scheme feasibility report hydrology—final draft
for comment. 03WAU/HYD-C, Riley Consultants, Takapuna, Auckland. 30 p. + appendix.
NIWA (National Institute of Water & Atmospheric Research) 2005: Atlas of New Zealand freshwater
fishes. www.niwa.cri.nz/rc/freshwater/fishatlas/species/dwarf galaxias.htm
Pascoe, A. 1995: The effects of vegetation removal on rabbits (Oryctolagus cuniculus) and small
mammalian predators on braided riverbeds of the Mackenzie Basin. Unpublished MSc. thesis,
University of Otago, Dunedin. 118 p.
Rebergen, A.; Keedwell, R.; Moller, H.; Maloney, R. 1998: Breeding success and predation at nests
of banded dotterel (Charadrius bicinctus) on braided riverbeds in the central South Island,
New Zealand. New Zealand Journal of Ecology 22: 33–41.
Ryder, G.; Kessing, V. 2005: Proposed Wairau hydroelectric power scheme: ecological investigations
of the Wairau River (Scope 2). Unpublished report prepared for Russell McVeagh on behalf
of TrustPower Ltd. 104 p.
Stokell, G. 1959: Notes on galaxiids and eleotrids with descriptions of new species. Transactions of
the Royal Society of New Zealand 87: 265–269.
Taylor, I.R.; Roe, E.L. 2004: Feeding ecology of little terns Sterna albifrons sinensis in south-eastern
Australia and the effects of pilchard mass mortality on breeding success and population size.
Marine and Freshwater Research 55: 799–808.
Trustpower 2005: Wairau Valley Hydroelectric Power Scheme, Volume 1—The assessment of
environmental effects report, June 2005. 167 p.
20
Hay—Dwarf galaxias habitat requirements